Sources & Categorization
Configure the content sources that feed the wiki, then run an AI pass to verify and refresh how docs are filed — reviewing every proposed Space + tag change before it is applied. Email & ClickUp ingestion is handled by the external MCP-push integration via /api/ingest; this page manages configuration and re-categorization.
Content sources
Registered sources that feed the library. Email & task content is pulled by an external push integration into the /api/ingest endpoint (plus a scheduled scan) — configuring a source here does not fetch on its own. Repo sources are informational pointers.
Categorization audit
Runs the AI classifier over indexed docs and proposes Space + tag changes where the current categorization drifts from the taxonomy. Nothing changes until you apply a proposal.
Content safety
Scans the docs you can read for material that shouldn’t be team-wide — secrets/keys, PII & HR data, personal contacts, offensive language. Config docs (env-var names) are ignored. Runs automatically on every ingest, too.